A dual-driven approach for refined modeling and performance analysis of heavy-duty gas turbine

Mechanism modeling of gas turbine relies on complex thermodynamic mechanisms and detailed component dimensions, often challenging to acquire. While data -driven modeling relies on operational data to establish the relationship between system inputs and outputs, accurately calculating unmonitored thermal parameters remains challenging. The dual -driven approach in gas turbine modeling integrates both thermodynamic mechanisms and actual operational data. Therefore, this study aims to establish a refined model for heavy-duty gas turbines by the dual -driven approach, and conduct its performance under hydrogen rich fuel. In order to consider the cooling effects, a correction coefficient is introduced for velocity loss coefficient of turbine's first -stage stator blade. Additionally, a surrogate model is built for the thermal calculation process of turbine first -stage stator blade. The results show that the heavy-duty gas turbine model has high accuracy. The calculation accuracy of the compressor pressure ratio and efficiency, turbine outlet pressure and electrical power reaches 0.965, 0.985, 0.965 and 0.95, respectively. Under the same fuel mass flow conditions as natural gas fuel at the design point, electrical power increases and overall efficiency decreases as the hydrogen volume content ranges from 0 to 18%. However, electrical power and overall efficiency decrease when the hydrogen volume fraction exceeds 18%. In all, the proposed modelling for heavy-duty gas turbine has good reliability and accuracy.